Use rotational burning
Overall effectiveness category Evidence not assessed
Number of studies: 17
Background information and definitions
Open habitats, such as grassland, heathland, peatland and fynbos, require disturbance processes to prevent them undergoing succession to scrub or woodland. One option for preventing succession and opening up areas of habitat is to use prescribed burning. However, burning is likely to be destructive in the short-term, because of either direct mortality or a reduction in food availability for butterflies and moths (Glaves et al. 2013). Rotational burning, where sites are burned once every few years, may allow time for habitat patches to recover to the optimal condition for butterflies and moths in the period between burns, while maintaining a dynamic patchwork of suitable habitat across a site. However, the impact of fires can vary depending on their exact characteristics, such as the frequency, temperature, ground surface intensity, time and size of burning compared to the surrounding unburned areas (Swengel 2001, Tucker 2003, New et al. 2010) and caution should be taken before instigating fire in place of alternative management options such as grazing or cutting.
For studies on the impact of single prescribed burns, see “Use prescribed fire to maintain or restore disturbance in grasslands or other open habitats”. For studies of other changes to the burning regime, see “Change season/timing of prescribed burning” and “Leave some areas unburned during prescribed burning”.
Glaves D.J., Morecroft M., Fitzgibbon C., Lepitt P., Owen M. & Phillips S. (2013) Natural England Review of Upland Evidence 2012 - The effects of managed burning on upland peatland biodiversity, carbon and water. Natural England Evidence Review, Number 004 (NEER004).
New T.R., Yen A.L., Sands D.P.A., Greenslade P., Neville P.J., York A. & Collett N.G. (2010) Planned fires and invertebrate conservation in South East Australia. Journal of Insect Conservation, 10, 567–574.
Swengel A.B. (2001) A literature review of insect responses to fire, compared to other conservation managements of open habitat. Biodiversity and Conservation, 10, 1141–1169.
Tucker G. (2003) Review of the impacts of heather and grassland burning in the uplands on soils, hydrology and biodiversity (ENRR550). Natural England (English Nature) report.
Supporting evidence from individual studies
A replicated, site comparison study in 1988–1993 in 51 tall-grass prairies in Illinois, Iowa, Minnesota and Wisconsin, USA (Swengel 1996) found that the abundance of prairie specialist butterflies in burned prairies increased with time since the last fire, but the abundance of other species was highest sooner after burning. At sites burned >4 years ago, the total abundance of six prairie specialists (75 individuals/hour) was higher than at sites burned the previous winter (10 individuals/hour). However, the abundance of grassland species and generalists was higher in the third year after burning (100–130 individuals/hour) than in the first or second year (50–100 individuals/hour) or fourth or fifth year (60–90 individuals/hour). The abundance of migrant species was higher in the year immediately after burning (770 individuals/hour) than in any subsequent year (10–40 individuals/hour). See paper for individual species results. Fifty-one sites (1–445 ha) were primarily managed by cool-season fire covering 5–99% of the site. From June–August 1988–1993, butterflies were surveyed 1–4 times/year on a transect through most sites (only Minnesota and Wisconsin in 1988–1989). Transects were sub-divided by the most recent management. Thirty-two species observed >49 times and at >5 sites were included, and divided into “prairie specialists” (6 species, only found on prairies), “grassland species” (13 species, found in prairies and other grasslands), “generalists” (10 species, found in grasslands and other habitats) and “migrants” (3 species, only present in the study area during the growing season).Study and other actions tested
A replicated, site comparison study in 1988–1996 on 17 upland prairies in Missouri, Minnesota, North Dakota and Wisconsin, USA (Swengel & Swengel 1997, same experimental set-up as Swengel 1998 and Swengel & Swengel 2001) found that prairies managed by rotational burning had a lower abundance of six out of seven specialist butterfly species than prairies managed by haying or grazing, or unmanaged areas. Of seven prairie specialist butterfly species, three (gray copper Lycaena dione, regal fritillary Speyeria idalia, arogos skipper Atrytone arogos) were less abundant in rotationally burned areas than in unmanaged areas, and four were less abundant in burned areas than in hayed (regal fritillary, Pawnee skipper Hesperia leonardus pawnee, Dakota skipper Hesperia dacotae) or grazed (Gorgone checkerspot Chlosyne gorgone) prairies. Poweshiek skipperling Oarisma poweshiek abundance in burned prairies was not significantly different from in hayed, grazed or unmanaged areas. See paper for individual species data. Across 17 prairies (16 to >120 ha), eight areas were managed by burning on rotation, six by haying (often in rotation), three by burning and haying, two by grazing, and two were unmanaged. From 1988–1996, butterflies were surveyed on transects through different management areas at each site. Sites were not surveyed in every year.Study and other actions tested
A replicated, site comparison study in 1988–1996 in 106 tallgrass prairies in Illinois, Iowa, Wisconsin, North Dakota, Minnesota and Missouri, USA (Swengel 1997) found that rotational burning, either alone or in combination with other managements, was associated with lower butterfly abundances for three of six fritillary butterfly species in at least one of the studied regions. In all regions where present, variegated fritillary Euptoieta claudia, great spangled fritillary Speyeria cybele and small pearl-bordered fritillary Boloria selene abundances were not affected by management type. Aphrodite fritillary Speyeria aphrodite abundance was lower in sites managed with burning or burning and mowing than in unmanaged and grazed sites in the Eastern Upper Midwest, but not in the Western Upper Midwest. Regal fritillary Speyeria idalia abundance was lower in sites with burning than grazing in the Upper Eastern Midwest, in sites with burning than haying in the Western Upper Midwest and in sites with burning or burning and haying than in sites with haying alone in southwestern Missouri. Meadow fritillary Boloria bellona abundance was lower in sites with burning or burning and mowing than in unmanaged sites in the Eastern Upper Midwest. All other management comparisons were not significantly different or had insufficient data for comparisons. Raw data were not provided. Butterfly surveys were conducted in June–September 1988–1996 in the Eastern Upper Midwest (northern Illinois, eastern Iowa and southwestern Wisconsin), Western Upper Midwest (eastern North Dakota, western Minnesota and western Iowa) and southwestern Missouri in tallgrass prairie sites with no management (7 sites), unknown management (4 sites) or managed by rotational burning (49 sites, mostly with a 2–5 year interval), rotational haying (24 sites, mostly with a 2–3 year interval), grazing (1 site) or a combination of burning and haying (20 sites) or burning and mowing (1 site). Not all sites were surveyed for the whole period or every year. Surveys were of varying lengths and conducted simultaneously along one set of parallel transects (5–10 m apart) in each site. Not all butterflies were found in all regions.Study and other actions tested
A replicated, site comparison study in 1986–1995 in 104 tallgrass prairies and 141 pine barrens in Illinois, Iowa, Minnesota, Missouri, North Dakota and Wisconsin, USA (Swengel 1998, same experimental set-up as Swengel & Swengel 1997 and 2001) found that rotationally burned grasslands or barrens had a higher abundance of two of 16 specialist butterfly species, but a lower abundance of seven specialist species, than either unmanaged sites or sites managed by grazing or mowing. Of 16 prairie or pine barren specialist butterfly species, two (ottoe skipper Hesperia ottoe and dusted skipper Atrytonopsis hianna) were more abundant in sites managed by rotational burning than grazed, mown or unmanaged sites. Seven species were less abundant in rotationally burned sites than in sites managed by haying (Dakota skipper Hesperia dacotae, pawnee skipper Hesperia leonardus pawnee), mowing (Persius duskywing Erynnis persius), cutting (cobweb skipper Hesperia metea), grazing and haying (arogos skipper Atrytone arogos, regal fritillary Speyeria idalia), or unmanaged sites (arogos skipper Atrytone arogos, Poweshiek skipper Oarisma poweshiek). Seven species had similar abundance between management types. See paper for individual species data. Of 104 prairies (1–2,024 ha), 61 were managed by cool-season burning on a 2–5-year rotation, of which 21 were additionally mown or hayed; 27 were hayed in summer on a 1–2-year rotation, of which two were also grazed occasionally with cattle; 10 were grazed; and six had not been managed for many years. Of 141 pine barrens, some were burned by wildfires, some were used for off-road vehicle trails, and some were power line rights-of-way (no further detail provided). From April–September 1986–1995, butterflies were surveyed on transects at each site. Most sites were surveyed more than once/year, and in >1 year.Study and other actions tested
A replicated, site comparison study in 1993–1997 in five oak savannas and prairies in Indiana, USA (Kwilosz & Knutson 1999) reported that sites burned on rotation had a similar abundance of Karner blue butterfly Lycaeides melissa samuelis to unburned sites. Results were not tested for statistical significance. Over 2–3 years, at two sites managed by rotational burning, the maximum number of Karner blue adults recorded increased from 159–288 to 296–725. Over 1–2 years, at two unburned sites, the abundance increased from 45–103 to 71–184. At a third unburned site, there were 104 adults in 1994 and 103 in 1997. Within a 6,000-ha reserve, two sites were each divided into four units. At a 177-ha site, one unit was burned annually in autumn or spring from 1993–1996, and adjacent units were not burned in consecutive years. At a 59-ha site, one unit/year was burned in autumn 1995–1996. Within the burned units, multiple 50–300-m2 patches with high Karner blue and wild lupine Lupinus perennis densities were left unburned. Three sites (areas not given) were not burned during the study. Butterflies were surveyed along a fixed 1.5–6.5-km transect/site, passing through all units. The highest number recorded at each site was taken as an annual population estimate. In July 1994, two surveys were conducted at each of two sites (one burned and one unburned). In June–August 1995–1997, three–nine surveys were conducted at each of 4–5 sites/year. Surveys were conducted 1–10 days apart.Study and other actions tested
A replicated, site comparison study in 1990–1997 in 86 tallgrass prairies and 32 pine-oak barrens in Illinois, Iowa, Minnesota, Missouri, North Dakota and Wisconsin, USA (Swengel & Swengel 2001, same experimental set-up as Swengel & Swengel 1997 and Swengel 1998) found that barrens had a higher abundance and species richness of specialist and grassland butterflies after more recent burning, but that prairies had a lower abundance and species richness of butterflies after more recent burning. In pine-oak barrens, the abundance and species richness of specialist and grassland butterflies was higher in recently burned areas than in unmanaged areas. However, in the first year after prairies had been burned, the abundance and species richness of specialist and grassland butterflies was lower than in areas which had not been burned in the last year. In addition, in Illinois, Wisconsin and eastern Iowa, the abundance of specialists, and the abundance and richness of grassland species, was lower in burned prairies of any age than in unmanaged prairies. However, apart from in the first year since burning, there were no differences between burned and unmanaged prairies in western Iowa, Minnesota and North Dakota. All data were presented as models results. A total of 77 prairies and 20 barrens were managed by rotational burning (every 2–5 years) in the cool-season, although some areas had not been burned for eight years. Nine prairies and 12 barrens had been unmanaged for many years. From May–September 1990–1997, butterflies were surveyed on parallel transects (5–10 m apart) at each site. Most sites were surveyed more than once/year, and in >1 year. Butterflies were classified as “specialists” (of native plants), “grassland” (occurring widely in open habitat) and “generalist” (occurring in a range of habitats).Study and other actions tested
A replicated study in 1998–1999 in three mistbelt grasslands in KwaZulu-Natal province, South Africa (Lu and Samways 2002) reported that populations of Karkloof blue butterfly Orachrysops ariadne persisted for at least a year following prescribed burning. In 1999, after prescribed fires in 1998, the authors recorded adults at Stirling, 90 hatched eggs at The Start and adults and 170 hatched eggs at Wahroonga. In March–June 1999, vegetation was searched for eggs at three mistbelt grasslands – Stirling (1–2 ha of Karkloff blue habitat), The Start (1 ha of habitat) and Wahroonga (10 ha of habitat).). Each egg was marked with a tag and a serial number so its hatching progress could be recorded. Details of adult butterfly recording were not provided. All three sites were in long term prescribed rotational burning regimes. Half of The Start was burned in July 1998, all of Wahroonga was burned in August 1998, and most of Stirling was burned in October 1998. No burning happened in 1999. Additionally, at The Start 50 Indigofera woodii H. Bol. var laxa saplings were planted in November 1998 and bramble was controlled with herbicide in January 1999. Stirling was cattle grazed each summer.Study and other actions tested
A replicated, paired, controlled study in 1992–1997 in 21 tallgrass prairie remnants in Illinois, Indiana and Wisconsin, USA (Panzer 2002) found that most insect populations (including butterflies and moths) which initially declined after burning recovered within a year and had higher abundance on burned than unburned sites. One year after burning, the abundance of 68% of insect populations which initially declined had recovered to match the abundance at unburned sites, and all recovered within two years. Moreover, 84% of these populations recovered to higher abundance in recently burned than unburned areas. Species which increased (26%) or did not change (34%) in abundance immediately after fire were not considered further. See paper for individual species results. Twenty-one prairies (2–600 ha) were divided into two or more units, and from 1992–1997 either 0 or 1 unit/prairie was burned each year in March–April. Recently burned units were left unburned for 2–3 years. Insects of 151 species were sampled in burned and unburned areas in a variety of ways. From May–October 1992–1997, random sweep samples were conducted at each site. In autumn 1992–1997, adhesive-coated plastic plates were placed to catch leafhoppers (Cicadellidae). Six moth caterpillars were sampled from 100–4,700 plant stems on 28 occasions. On three nights in 1997, black-light traps were used at two sites to sample three moth species. From June–July 1995–1997, butterflies were surveyed along transects and on 5-minute counts from random points at seven sites.Study and other actions tested
A replicated, site comparison study in 2004–2005 in two remnant prairies and adjacent land in Iowa, USA (Vogel et al. 2007, same experimental set-up as Vogel et al. 2010) found that prairies which were rotationally burned and grazed had a higher abundance, but not species richness, of butterflies than prairies which only received burning or grazing, but the three management practices supported different species. The abundance of butterflies in burned and grazed prairies (31.5 individuals/unit) was higher than in prairies which were only burned (20.2 individuals/unit) or only grazed (27.8 individuals/unit). Species richness of butterflies was similar in prairies managed by burning and grazing (8.5 species/unit), only burning (8.6 species/unit) and only grazing (7.4 species/unit). Butterfly diversity was higher in prairies managed by burning only than in prairies managed by grazing only or grazing and burning (data presented as model results). However, each management practice supported different species (see paper for details). Across two remnant prairie reserves (320 and 1,800 ha) and surrounding land, 28 management units (10–167 ha) were managed consistently for ≥4 years. Ten units were burned during autumn or spring every 1–6 years. Six units were lightly grazed on rotation (1 cow-calf pair/4 ha). Twelve units were burned and grazed. From June–August 2004–2005, butterflies were surveyed for 30 minutes twice/year at 69 sites (50 × 50 m, >150 m apart) across the 28 units.Study and other actions tested
A replicated, site comparison study in 2004–2005 in two remnant prairies and adjacent land in Iowa, USA (Vogel et al. 2010, same experimental set-up as Vogel et al. 2007) found that recently burned sites had lower abundance and species richness of butterflies than sites burned longer ago. At sites burned 7–25 months before surveying, the abundance and species richness of all butterflies (abundance: 6–21 individuals/unit; richness: 4–7 species/unit) and of prairie specialists (abundance: 2–14 individuals/unit; richness: 1–3 species/unit) were lower than at sites burned 43–61 months before surveying (all: 13–21 individuals/unit; 5–7 species/unit; specialists: 5–15 individuals/unit; 2–3 species/unit). Of 14 individual species, four were less abundant, and three were more abundant, at recently burned sites than at sites burned longer ago (see paper for details). Across two remnant prairie reserves (320 and 1,800 ha) and surrounding land, 22 management units (10–102 ha) were managed consistently for ≥4 years. All units were burned during autumn or spring every 1–6 years. Twelve units were also lightly grazed on rotation (0.25 cow-calf pairs/ha). From June–August 2004–2005, butterflies were surveyed for 30 minutes twice/year at 53 sites (50 × 50 m, >150 m apart) across the 22 units. Butterflies were classified as “specialists” which require native prairie plants as adults or caterpillars, and “generalists” which use a variety of common plants and occur in a range of open habitats.Study and other actions tested
A replicated, before-and-after, site comparison study in 1979–2009 in prairie, grassland and pine barrens in Illinois, Iowa, Minnesota and Wisconsin, USA (Swengel et al 2011) found that rotational burning, along with other interventions, had mixed effects on prairie-specialist butterfly abundance. At sites managed by rotational burning, the Ottoe skipper Hesperia ottoe declined over time in some sites in two of four states (Minnesota and Wisconsin), and the Poweshiek skipperling Oarisma poweshiek and Aphrodite fritillary Speyeria Aphrodite declined in sites in one state (Iowa). Regal fritillary Speyeria idalia numbers increased at one site in one state (Wisconsin). No changes were found for species monitored in other states. Results are presented as statistical tests. In four areas in Wisconsin designated as butterfly reserves (≤32% burned/year), Karner blue butterfly Lycaeides melissa samuelis abundance decreased less (1999: 27 individuals/km; 2009: 20 individuals/km) than at 10 sites where higher levels of burning, as well as timber extraction and public rights-of-way, were allowed (1999: 17 individuals/km; 2009: 2–5 individuals/km). Butterfly counts from 10 separate researcher teams, which overlapped in location, were combined and adjusted to account for differences in survey methods. At each site, 20–35% of the area was burned on average each year, with each section being burned at least every 10 years. Authors reported that sites were also managed with mowing, haying, brush-cutting and herbicides; details of management at each not provided.Study and other actions tested
A replicated, randomized, controlled study in 2007–2009 in 13 tallgrass prairies in Iowa and Missouri, USA (Moranz et al 2012) found that butterfly species richness did not differ between pastures with rotational burning and grazing and pastures where the whole area was burned (with or without grazing), but the density of some species differed with management and year. There was no difference in species richness between pastures burned rotationally with grazing (6–7) and pastures with complete burning and grazing (6–7) or no grazing (7). Regal fritillary Speyeria idalia densities were lower with rotational burning with grazing and complete burning with grazing (both average: 0 individuals/ha) than with complete burning and no grazing (5–7 individuals/ha). The same was seen in the common wood-nymph Cercyonis pegala (rotational burning/grazing or complete burning/grazing = 2 individuals/ha; complete burning/no grazing = 21) and the monarch butterfly Danaus plexippus (rotational burning/grazing or complete burning/grazing = 1 individuals/ha; complete burning/no grazing = 2) but only in 2009 (the survey after the burn in non-rotational pastures, no difference in 2008). In 2008 only, eastern tailed-blue Cupido comyntas population densities were higher in rotational burning with grazing pastures (124 individuals/ha) than complete burning with (101) or without grazing (55) pastures (no difference in 2009). Other species analysed were not affected by management in either year. Pastures (15–31 ha) were managed in one of three ways: 1) rotational burning, with a third of each pasture burned annually in spring 2007–2009 and grazing May–October annually (five pastures), 2) complete burning in spring 2008 or 2009 and grazing May–October annually (four pastures), or 3) complete burning in spring 2008 or 2009 without grazing (four pastures). Butterflies were surveyed in June–July 2008–2009 along six 100-m transects/pasture.Study and other actions tested
A replicated, randomized, paired, controlled study in 2005–2007 in four tallgrass prairies in Missouri, USA (Moranz et al. 2014) found that spring burning reduced regal fritillary Speyeria idalia abundance in mid-summer, but increased abundance later in the summer. In late June, the number of regal fritillaries in plots burned that spring (3–12 individuals/ha) was lower than in plots burned one (12–22 individuals/ha) or two (7–25 individuals/ha) years earlier. However, by late July, the number was higher in recently burned plots (14–22 individuals/ha) than in plots burned one or two years earlier (1–9 individuals/ha). In early June, fritillary abundance was not significantly different in recently burned plots (5–22 individuals/ha) and those burned in previous years (7–9 individuals/ha). From 2000–2004, four remnant prairies were burned on rotation and occasionally hayed or lightly grazed. In 2005, half of each prairie was randomly assigned to one of two treatments: grazing and rotational burning, or rotational burning only. Each half was sub-divided into three plots (20–34 ha), which were randomly assigned to be burned in either March 2005, 2006 or 2007. The grazed sites were stocked with cattle (2.2 ha/animal unit) annually from April–August. From June–July 2006–2007, butterflies were surveyed three times/year on a transect through each plot.Study and other actions tested
A replicated, site comparison study in 1988–2013 in 24 prairies, pine-oak barrens and fields in Wisconsin, USA (Swengel & Swengel 2015) found that prairies managed by rotational burning had more strongly declining populations of grass-skipper butterflies (Hesperiinae) than unmanaged pine barrens or lightly managed fields. In prairies managed by rotational burning, specialist and grassland skipper butterflies had more strongly declining population trends than in pine-oak barrens (specialist and grassland species) or old fields (grassland species only) (data presented as model results). Ten native prairies were managed by cool-season burning, typically on a 2–5-year rotation, with some mowing, cutting and spot herbicide application. Eight pine-oak barrens were last burned by wildfires in 1977 or 1988. Six fields reverting from agriculture were managed by burning, grazing, haying, cutting, herbicide or tilling, but with no more than 10% of a site managed each year. In spring and summer 1988–2013, butterflies were surveyed along transects (2–3 km/hour) at each site, generally at least twice/year and in at least two years/site. Skippers were classified as eight “specialist” (restricted to native grassland habitat) or five “grassland” (occurring widely in open habitat) species.Study and other actions tested
A site comparison study in 2014 in a restored grassland and oak barren landscape in Indiana, USA (Shuey et al. 2016) reported that regal fritillary Speyeria idalia were found across a landscape restored by planting and managed by rotational burning. Results were not tested for statistical significance. Eighteen years after planting and rotational burning began, on four restoration sites with high plant diversity, the abundance of regal fritillaries peaked at 17 butterflies/30-minute transect, compared to 12 butterflies/transect on two remnant prairies and a low plant diversity restoration site, 19 butterflies/transect in an old field, and 0 butterflies/transect in an agricultural field. Prior to restoration, authors reported that regal fritillaries were only found at three small sites in the landscape. Beginning in 1996, over 3,240 ha of agricultural land was restored to native grassland and oak barrens by planting seed mixes containing over 620 native species. In addition, seeds and plugs of arrowleaf violet Viola sagittata and bird's-foot violet Viola pedata were planted as host plants. The area was managed to control invasive species and, once established, patches were burned on a three-year rotation. From May–September 2014, butterflies were surveyed every two weeks on 30-minute transects at nine sites across the landscape: four restoration sites with high plant diversity, one restoration site with low plant diversity, two remnant prairies, one old field, and one site still in agricultural production, none of which had been burned during the previous year.Study and other actions tested
A replicated, site comparison study in 2012–2013 in 12 semi-natural grasslands in Nagano Prefecture, Japan (Uchida et al. 2016) found that meadows managed by traditional rotational burning and mowing had a higher species richness and diversity of butterflies than annually burned, annually mown or abandoned meadows. In rotationally managed meadows, the diversity and species richness of threatened (6–7 species/meadow) and common (10–12 species/meadow) butterflies was higher than in annually burned (threatened: 2–3; common: 6 species/meadow), annually mown (threatened: 3; common: 4 species/meadow) or abandoned meadows (threatened: 1–2; common: 1–2 species/meadow) (diversity data presented as model results). Three meadows were managed traditionally: each year half of the meadow was burned in April and mown in September, while the other half was unmanaged, and management rotated each year. An additional three meadows had been burned annually for 7–13 years, three meadows had been mown annually in April or August for 8–9 years and three meadows had been abandoned (unmanaged) for 6–13 years. From May–September 2012–2013, butterflies were surveyed monthly on three 5 × 30 m plots/meadow.Study and other actions tested
A replicated, site comparison study in 1997–2016 in seven tallgrass prairies in Wisconsin, USA (Henderson et al. 2018) found that burning initially reduced the abundance of regal fritillaries Speyeria idalia but then resulted in increased abundance for up to four years. All data were presented as model results. Regal fritillary abundance was reduced immediately after burning, highest four years after burning, and reduced again by eight years after burning. Areas burned within the last 7 years had a higher abundance of regal fritillaries than unburned areas, but fritillary abundance was not significantly higher in more frequently burned areas. The proportion of habitat burned did not affect abundance. From 1997–2016, seven patches of remnant and restored prairie (19–41 ha, 0.25–3.5 km apart) were managed with periodic rotational burning, where 25–93% of each site was left unburned each year. Burned areas always had suitable unburned habitat within 622 m. In July and August 1997–2016, beginning 7–10 days after the first emergence, regal fritillaries were surveyed weekly on 57 permanent transects across the seven sites, with >3 surveys/transect/year. The maximum number of regal fritillaries recorded on a single survey each year was used as the population estimate for each transect.Study and other actions tested